Sathish SathishN

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

When to Use Each Version

claudette-mimir.md ⭐ NEW!!!

✅ Full integration with NornicDB's memory bank
✅ Graph functions enable multi-hop reasoning

CLICK HERE-> Claudette NornicDB

NornicDB Memory Bank

GitHub Copilot Custom Agent Mode

I modified that beast mode to be more robust, and I love it. Refactored an entire older repo I made from years ago just for fun and was able to do perfectly

I call it "Extensive Mode"

"Extensive Mode.chatmode.md"

The tools should be customized per environment and what you have available,

description

4.1 Beast Mode

tools

changes

codebase

editFiles

extensions

fetch

findTestFiles

githubRepo

new

openSimpleBrowser

problems

readCellOutput

runCommands

runNotebooks

runTasks

runTests

search

searchResults

terminalLastCommand

terminalSelection

testFailure

updateUserPreferences

usages

vscodeAPI

Beast Mode v3 is out now: 👉 https://gist.github.com/burkeholland/88af0249c4b6aff3820bf37898c8bacf

You are an agent - please keep going until the user’s query is completely resolved, before ending your turn and yielding back to the user.

🗺️ OSM - self host the entire planet 🌎 in ~30 minutes 🚀

TL;DR

mkdir osm
wget -O osm/planet.mbtiles https://hidrive.ionos.com/api/sharelink/download?id=SYEgScrRe
podman run -ti --rm -p 9000:9000 --name sms -v $(pwd)/osm/:/data/ registry.gitlab.com/markuman/sms:latest
firefox http://localhost:9000

UniFi Network Application on MikroTik RB5009

Self-hosted UniFi Network controller and MongoDB running as two RouterOS containers on a MikroTik RB5009UPr+S+, backed by a USB stick split into swap and ext4 data partitions.

Hardware

Router: MikroTik RB5009UPr+S+
- CPU: Marvell Armada A72 (Cortex-A72), ARMv8.0-A, 4 cores @ 1.4 GHz
- RAM: 1024 MiB
RouterOS: 7.22.3 (stable)

	"""
	The most atomic way to train and run inference for a GPT in pure, dependency-free Python.
	This file is the complete algorithm.
	Everything else is just efficiency.

	@karpathy
	"""

	import os # os.path.exists
	import math # math.log, math.exp

	from __future__ import annotations

	import argparse
	import datetime as dt
	import json
	import math
	import re
	import statistics
	from dataclasses import dataclass
	from pathlib import Path